Dependence of UGT76B1-mediated response on SA and JA pathway

Previously, UGT76B1 has been shown to be a regulator in SA-JA cross-talk, suppressing SA pathway and activating JA pathway (von Saint Paul, 2010).

To elucidate how UGT76B1 integrates into SA and JA pathways, both ugt76b1-1 and UGT76B1-OE-7 were introgressed in lines deficient in either SA or JA pathways. The NahG overexpression lines completely block SA accumulation by introducing a bacterial NahG gene, encoding a hydrolase activity towards SA (Gaffney et al., 1993). The sid2 mutant leads to the loss of stress-induced SA synthesis, where only a basal level of 5-10% SA remains (Nawrath and Metraux, 1999). The npr1 mutation causes the loss-of-function of a major positive regulator activating the SA pathway through interaction with TGA transcription factors (Cao et al., 1997). The mutant jar1 blocks JA pathway due to inability to synthesize bioactive JA-Ile conjugates (Berger, 2002).

Marker genes of SA or JA pathways were measured by RT-qPCR to evaluate SA or JA-mediated response. Marker genes for SA pathway include PR1, EDS1 (Figure 1), SAG13, and WRKY70 (Figure 3), while marker genes for JA pathway are PDF1.2 and VSP2 (Figure 2).

The induction of PR1 and SAG13 by ugt76b1-1 was completely abolished in ugt76b1 NahG with the same background expression level as in NahG, suggesting that induction of both genes in ugt76b1 was dependent on SA levels. However, PR1 and SAG13 can still be effectively induced by ugt76b1 in a sid2 background, independent from SID2. Overexpression of UGT76B1 was still effective to suppress expression of PR1 and SAG13 in sid2 indicating that the suppression of PR1 and SAG13 expression by UGT76B1 overexpression was independent from SID2 (Figure 7A). On the other hand, the expression of the JA marker VSP2 in ugt76b1 NahG was enhanced to the same level as in NahG, indicating that the suppression of the JA pathway by ugt76b1-1 was dependent on SA levels. UGT76B1 was effective to induce VSP2 expression in a sid2 background, suggesting that the UGT76B1-dependent enhancement of VSP2 was independent from SID2 (Figure 7A). However, the loss-of-function of UGT76B1 could not further suppress VSP2 expression in sid2. Notably, VSP2 was only suppressed by the ugt76b1 mutant around twofold, much less than the change of more than tenfold induction by UGT76B1 overexpression. This suggested that VSP2 was less sensitive to the regulation by ugt76b1 than to the regulation by UGT76B1 overexpression. Since basal

UGT76B1 expression was quite low in Col-0 whereas UGT76B1 overexpression greatly increased the expression of UGT76B1, the different sensitivities of VSP2 in response to the ugt761 knockout and the UGT76B1 overexpression could be explained. Thus, the lack of regulation of VSP2 in response to loss-of-function of UGT76B1 in sid2 could be also due to the lower sensitivity. Therefore the positive regulation of VSP2 by UGT76B1 might be independent from SID2.

The activation of PR1 and SAG13 by ugt76b1 was still effective in npr1 (Figure 7B), suggesting an NPR1-independent manner of induction. In agreement, overexpression of UGT76B1 was also effective to suppress PR1 and SAG13. Thus, UGT76B1 could regulate expression of PR1 and SAG13 independent from NPR1. The suppression of EDS1 and WRKY70 by UGT76B1 overexpression was also effective in npr1, suggesting a regulation of EDS1 and WKRY70 independent from NPR1. In contrast, the expression of EDS1 and WRKY70 in ugt76b1 npr1 was the same as in npr1. Though the loss-of-function of UGT76B1 seemed not to regulate EDS1 and WRKY70 in npr1, it at least did not contradict the effect of UGT76B1 overexpression. EDS1 and WRKY70 were only slightly induced in the ugt76b1 knockout whereas they were much stronger suppressed by UGT76B1 overexpression as compared to Col-0 (Figure 7B). This suggested that the regulation of EDS1 and WKRY70 was not sensitive to the loss-of-function of UGT76B1, probably because the basal expression of UGT76B1 in Col-0 was already very low. The UGT76B1 overexpressor greatly increased UGT76B1 expression and activated EDS1 and WRKY70 effectively. Therefore, UGT76B1 mostly likely negatively regulated EDS1 and WRKY70 independent from NPR1 (Figure 7B).

The constitutive expression of UGT76B1 effectively increased VSP2 expression in the npr1 background suggesting that the UGT76B1-dependent enhancement of VSP2 was independent from NPR1. In contrast, the loss-of-function of UGT76B1 seemed to induce VSP2 in npr1 also, suggesting the negative regulation of VSP2 expression by UGT76B1 in npr1 background, contrary to the positive regulation of VSP2 expression by UGT76B1 in Col-0.

This might suggest that the loss-of-function of UGT76B1 had a direct negative role in regulating VSP2 in the npr1 background whereas the positive regulation of UGT76B1 on VSP2 in Col-0 was mostly coming from the SA pathway via SA and JA cross-talk.

The jar1 mutant did not influence expression of PR1 and SAG13 in ugt76b1-1. The jar1 mutant compromised in the JA pathway probably resulted in reduction of suppression on the SA pathway by the JA pathway. It is reasonable that in the jar1 mutant, SA-mediated responses (PR1 and SAG13) could be activated relative to Col-0. However, PR1 and SAG13

were still suppressed by UGT76B1-OE-7 in combination with jar1 (Figure 7C). Therefore the suppression of PR1 and SAG13 by UGT76B1 was directly due to the SA pathway but not the antagonism of the JA pathway via cross-talk. The enhancement of VSP2 in UGT76B1-OE-7 was eliminated by jar1, suggesting the activation of JA response by UGT76B1-OE-7 was dependent on JAR1.

Thus, both the enhancement of SA and suppression of JA pathways by ugt76b1-1 were dependent on SA. UGT76B1 might regulate SA marker genes (PR1, SAG13, EDS1 and WRKY70) and JA maker gene (VSP2) independent from both SID2 and NPR1.

SAG13 is a senescence marker gene and at least partially SA-dependent (Miller et al., 1999;

Yoshida et al., 2001). Both the early senescence in ugt76b1-1 and delayed senescence in UGT76B1-0E-7 have been reported to be consistent with induction of SAG13 in ugt76b1-1 and suppression of SAG13 in UGT76B1-OE-7 relative to Col-0 (von Saint Paul, 2010). The expression of SAR13 was lower in both the ugt76b1 sid2 and ugt76b1 npr1 double mutants than in Col-0 (Figure 7A and B). In contrast, SAG13 expression in ugt76b1-1 was not influenced by the mutation of JAR1 at all (Figure 7C). This suggested that senescence marker gene SAG13 expression was dependent on SA and NPR1, but not JAR1. To investigate whether an early senescence phenotype developed in ugt76b1-1 was dependent on SA and JA pathways, I observed the senescence development in the ugt76b1 sid2, ugt76b1 npr1, ugt76b1 NahG and ugt76b1 jar1 double mutants. The early senescence phenotype was abolished in the ugt76b1 sid2, ugt76b1 npr1 and ugt76b1 NahG double mutants, suggesting that the early senescence developed in ugt76b1-1 was dependent on SA and NPR1. However the double mutant ugt76b1 jar1 still developed the early senescence phenotype as ugt76b1-1, indicating the early senescence phenotype of ugt76b1 was independent from JAR1. Interestingly, the retarded growth of the jar1 mutant was gone in both ugt76b1 jar1 and UGT76B1-OE-7 jar1.

No obvious growth difference was observed in UGT76B1-OE-7 with or without combination with jar1, sid2 and npr1 (Figure 8).

**

**

**

**

ugt76b1-1 UGT76B1-OE-7 sid2

NahG

ugt76b1-1 sid2 ugt76b1-1 NahG UGT76B1-OE-7 sid2

A

normalized expression vs. wt (Col-0) log10 -5 -4 -3 -2 -1 0 1 2 3

C

UGT76B1-OE-7 ugt76b1-1 jar1 UGT76B1-OE-7 jar1

normalized expressionvs. wt (Col-0) log10

PR1 SAG13 VSP2

**

2 **

-4 -3 -2 -1 0 1

B

-5 -4 -3 -2 -1 0 1 2

ugt76b1-1 UGT76B1-OE-7 npr1

ugt76b1-1 npr1 UGT76B1-OE-7 npr1

normalized expression vs. wt (Col-0) log10

*

*

**

*

** **

Figure 7. SA and JA marker genes expression in UGT76B1 overexpression and knockout lines after introgression into sid2, NahG, npr1 and jar1.

Gene expression of PR1, SAG13, VSP2, EDS1 and WRKY70 in 4-week-old ugt76b1-1, UGT76B1-OE-7 and double mutants (with: (A) sid2 and NahG, (B) npr1, (C) jar1) was measured by RT-qPCR.

Expression levels were normalized to UBIQUITIN5 and S16 transcripts; levels relative to Col-0 plants are displayed. Arithmetic means and standard errors from log10-transformed data of at least 4 independent replicates from two separate experiments were calculated using ANOVA. The significance for the difference of PR1, SAG13, VSP2, EDS1 and WRKY70 expression was measured, compared between lines (ugt76b1 NahG, UGT76B1 NahG, ugt76b1 sid2, UGT76B1 SID2, ugt76b1 npr1 and UGT76B1 npr1) and corresponding single mutants sid2, NahG (A), npr1 (B) and jar1 (C) respectively. Stars indicated the significance of the difference between the two bars connected by the dotted line: ** p-value < 0.01, * p-value < 0.05.

Col-0 jar1 sid2 npr1

ugt76b1-1

UGT76B1-OE-7

UGT76B1-OE-7 x sid2 ugt76b1-1 x sid2

UGT76B1-OE-7 x jar1

ugt76b1-1 x npr1

UGT76B1-OE-7 x npr1 ugt76b1-1 x jar1

Figure 8. The impact of UGT76B1 expression on the onset of senescence is dependent on SID2 and NPR1, but independent from JAR1.

Pictures were taken from four-week-old Arabidopsis plants.